The discovery of the natural products collectively known as the combretastatins from a willow tree (Combretum caffrum) in South Africa ushered in a new era in the development of antimitotic agents which inhibit the assembly of tubulin into microtubules. Combretastatin A-4 (CA-4) and combretastatin A-1 (CA-1), which have the structures: are especially potent in terms of in vitro cytotoxicity against human cancer cell lines and in their ability to inhibit the assembly of tubulin into microtubules through a direct interaction at the colchicine binding site on β-tubulin.
It is interesting and instructive to note that while both CA-4 and CA-1 are potent inhibitors of tubulin assembly and are strongly cytotoxic against human cancer cell lines (Table 1), both of these in vitro assays suggest that CA-4 is more active biologically than CA-1.
TABLE 1In Vitro Evalutation of Combretastatins and Combretastatin ProdrugsInhibition of TubulinMTT CytotoxicityMTT CytotoxicityPolymerization (IC50)(IC50) at 1 hour(IC50) at 5 hoursCA-41-2 uM0.1 uM0.O5 uMCA-12-4 uM 10 uM0.05 uMCA-4P>40 uM0.8 uM0.002 uM CA-1P>40 uM3.2 uM0.0046 uM 
However, when both of these analogs are converted to their corresponding prodrug forms (CA-4P and CA-1P accordingly) and evaluated in vivo in terms of tumor vascular shut-down (FIG. 1) and tumor growth delay (FIG. 2), then it is apparent that CA-1P is eight to ten-fold more active than CA-4P in SCID mice. CA4P and CA1P have the structures: 
In the case of CA-1P, the most probable biological mode of action ultimately appears to be an enzymatic cleavage by non-specific alkaline phosphatase (or a related enzyme) converting CA-1P (which is not active with tubulin) to the parent CA-1 (which is active with tubulin). CA-1 inhibits the assembly of cytoskeletal tubulin into microtubules resulting in a morphological change in the endothelial cells lining the microvessels of tumors. This morphological change causes the endothelial cells to “round-up” which results in an inability of the microvessels to sustain blood flow. Blood clotting and other events ensue which ultimately result in death of the surrounding tumor tissue. Healthy tissues are, for the most part, not affected even though the compound is administered systemically. Several possibilities exist for this selectivity including (but not limited to): (a) the possibility that there is enhanced activity or expression of nonspecific alkaline phosphatase in the micro-environment of the endothelial cells lining the tumor microvessels; (b) potential differences in the tubulin itself between mature healthy cells and immature, rapidly proliferating endothelial cells in the tumor microvessels which cause enhanced disruption of the tubulin assembly/disassembly process in the tumor microenvironment; (c) tumor cells are known to have “leaky” vessels and it is possible that some of the improved tumor growth delay is due to the compound (as parent drug or prodrug) leaving the blood vessels and entering the cytosol around the tumor where it can form a “supply pool” which ultimately enters the tumor cell itself and (as the parent compound) functions as an antimitotic agent inhibiting cellular division during metaphase of the cell cycle. The enhanced (10 fold) activity in vivo of CA-1P may be due, in part, to the pharmacokinetics associated with the cleavage of both of the phosphate groups (perhaps one cleaves more rapidly than the other) and the subsequent interaction of the parent diphenol (or perhaps one, or both, of the monophenols/monophosphates) with tubulin.
It has therefore been an object of the studies which led to the present invention to demonstrate and confirm that the enhanced activity of CA-1P may not be due entirely to the substitution pattern in the B-ring of 2,3-diphosphate salt, but rather may be due to a change in pharmacokinetics associated with a Z-stilbenoid compound which incorporates a 3,4,5-trimethoxyphenyl motif in the A-ring, and a 4′-methoxy, 3′-O-Phosphate, along with the incorporation of an additional group (with either an electronic or steric bias) at C-2′, C-5′, or C-6′. Compounds of this basic structural pattern may demonstrate good bioavailability and favorable pharmacokinetics which result in an improved interaction with tubulin and enhanced efficacy as VTAs. It should be readily apparent to anyone skilled in the art that although the new compounds described herein have a trimethoxyaryl substitution pattern in the A-ring, it is a logical extension to vary the positions of these methoxy groups in the C-2, C-3, C-4, C-5, and C-6 positions. Substitution patterns of this type may also result in compounds active as VTAs.
A variety of studies have suggested that the 3,4,5-trimethoxy substitution pattern on the A-ring and the 4-methoxy moiety on the B-ring are important structural features of the pharmacophore for these stilbenoid analogs (FIG. 3). Accordingly, the inventors have maintained these functionalities in most of the new molecules and have included further substitution patterns around the B-ring. The present invention and the compounds which are a part thereof is not limited in this respect, however, and substitution by other than a 4-methoxy moiety on the B-ring is contemplated. It is the contention of the present inventors that the improved in vivo activity of CA-1P (as related to CA-4P) is not due solely to the presence of a diphosphate moiety, but rather may have a strong tie to the pharmacokinetics of this compound including the enzymatic cleavage of the phosphate group (presumably by nonspecific alkaline phosphatase), subsequent inhibition of tubulin assembly resulting in morphological changes (rounding-up) of the immature endothelial cells lining the microvessels of tumors, and the resulting inability of these microvessels to sustain blood flow. Additional pharmacokinetic parameters such as reversibility of tubulin binding and perhaps incorporation of the parent CA-1 in the cytosolic fluid around the tumor cell itself may also play key biological roles.